Fig 2: Hyperosmolar sucrose induces the transcription of zonula occludens (ZO) protein –1 at mRNA but not at protein level. (A) qPCR analyses of n = 3 experiments showed a significantly increased ZO–1 mRNA expression after sucrose-treatment (Suc) as compared to controls (PBS). Shown are means (+SEM). Statistical significance: * p < 0.05. (B) Immunoblotting with antibodies against ZO–1 showed no change in protein expression after sucrose-treatment. Immunolabeling with antibodies against GAPDH served as loading control. Original images are shown in Supplementary File S1. (C) Quantification of immunoblots (n = 4) confirmed the constant level of protein expression. (D) Quantification of immunofluorescent maxima in images of Cx43 and ZO–1 immunofluorescence. Whereas the fluorescence maxima at 488 nm and 568 nm excitation wave lengths were close-by in PBS-treated cultures, peaks were more distant in sucrose-treated cells, indicating less co-localization of the two detected proteins. Quantification of the peak distances (in nm) are shown as box plots with median and whiskers for minimum and maximum; ** indicate p = 0.0033). (E) The first panel on the left shows the merged immunofluorescence color images of Cx43 (green) and ZO–1 (red) proteins in PBS– and sucrose-treated astrocytes. Nuclei were labeled by DAPI (blue). Monochrome images show the immunofluorescence of Cx43 (second panel) and ZO–1 (third panel) as well as DAPI (fourth panel) separately. Although cell membranes and cellular contact sites are labeled by Cx43 and ZO–1 under both conditions, the distribution of the two immunosignals seems to change upon sucrose treatment. Scale bar: 20 µm.

Fig 3: Sucrose does not affect Cx43 mRNA or protein expression. Astrocytes were treated with sucrose (n = 3) or with PBS as control (n = 3) for five minutes. (A) qPCR was performed with specific primers for Cx43 and results were normalized to 18S rRNA expression. Calculating the relative expression changes with respect to PBS-treated cells (=1) according to Pfaffl et al. revealed no changes. (B) SDS-PAGE of cell extracts and immunoblotting with an antibody recognizing all phosphorylation forms of Cx43 with GAPDH as a loading control were performed. Cx43 immunosignal showed unphosphorylated (P0) isoforms as well as two phosphorylated (P1/P2) isoforms. Original images are shown in Supplementary File S1. (C) Quantification of signal intensities, performed on the overall immunosignal per sample, did not show any significant changes of total Cx43 protein content. Shown are means + SEM of signal intensities relative to the control. (D) Differential quantification of immunopositive bands P0, P1, and P2 of PBS- (black columns) and sucrose- (gray columns) treated samples. Individual comparisons of P0, P1, and P2 band intensities between groups revealed significant differences for P0 bands (** indicates p < 0.0051; uncorrected Fisher’s LSD).

Fig 4: Cx43 protein expression in the Mfn-DKO versus WT hearts. (A) Representative immunoblots (cropped) of total Cx43 and phosphorylated Cx43 from 3 different WT versus Mfn-DKO hearts. Original gel images are presented in Supplementary Fig. 2. (B) Quantification of band intensity for total Cx43, normalized to GAPDH expression, *p < 0.05 when compared against WT. (C) Quantification of band intensity for phosphorylated Cx43, normalized to total Cx43.

Fig 5: Lateralization of Cx43 in the Mfn-DKO hearts. Cross-sections of left ventricle from 8-weeks-old males, (A) WT control versus (B) Mfn-DKO, cardiac-specific, transgenic mice, immuno-stained with anti-Cx43 (brown). Examples of lateralised regions of Cx43 are delineated with red arrows. The negative controls using normal rabbit IgG and secondary antibody solely, without primary antibody, are represented by (C,D) respectively. Scale bar = 20 µm (E) Quantitative analysis of Cx43 lateralization. Images are representatives of 3 hearts for each group. Data are presented as mean ± SEM; *p < 0.05 when compared against WT.